Splunk for Facebook … cont’d

To piggyback from the previous article, the purpose of this app is to help users make sense of the data retrieved from the Facebook Graph API. Despite the fact that we at Splunk love to sift through and look at huge amount of data, we still like to explore if we can cleanse them, in other words, massage, filter, extract important details, transform them in order to make them more meaningful and sensible.

Some interesting use cases that this app addresses:

[1] Correlating your connections

This app will ideally help the users, at a quick glance, understand how they are connected to other users within the Facebook ecosystem. We are talking about the interesting “n-th degree of separation” problem in the engineering domain that can be applied into our everyday lives. Remember the graph theory in our CS syllabus? The users (the “Facebookians”) will be represented as nodes in the graph. We are talking about how a set of users are mutually connected to other set of users. You might be surprised to realize how close you are connected to other users.

[2] Realtime feed updates

One of Splunk’s core strength is its realtime search and indexing. This app is envisioned to consume the real time status updates from users and then leveraging Splunk’s realtime search feature to look for interesting events that take place. How about scanning for a particular set of phrases or keywords? The possibilities are endless.

[3] Making sense of your data

Again, it’s all about deriving something into something useful. An example would be, providing users interesting statistics of the time spent on Facebook, the number of “likes” for the past 1 hour, the location of “checkins” for the past 3 days, to understand the latest change of preference such as from “Gucci” to “Prada” and probably the buzzwords for the particular period of time.

The possibilities are limitless …

Check out the latest app release at

Comments and feedback are very much welcomed. Now that the pandora box is opened, it’s the time now to look at the wealth of information it offers.

Until the next article …

Disclaimer: Opinions expressed in this blog do not represent the views of Splunk.

Nicholas Key

Posted by